Author's Personal Copy

Author's personal copy

J Insect Conserv DOI 10.1007/s10841-014-9649-1

ORIGINALPAPER

A spider diversity model for the Caucasus Ecoregion

Giorgi Chaladze • Stefan Otto • Sebastian Tramp

Received: 27 June 2013 / Accepted: 16 June 2014 � Springer International Publishing Switzerland 2014

Abstract Precise information on spatial patterns of spe- invertebrate taxa in the Caucasus Ecoregion is needed to cies richness and endemic species distribution is important improve conservation efforts in this biodiversity hotspot. for effective species conservation. In the Caucasus Ecore- gion such information is virtually non-existent for inver- Keywords Araneae� Biodiversity� Climatic variables� tebrate taxa. Using occurrence data from a large database Spatial patterns� Altitudinal gradient� Caucasus we calculated species distribution models with the GARP Ecoregion� Global hotspots algorithm for 471 spider species to visualize the diversity distribution of spider species in this region. Overall species diversity was highest in mountain forests of the North Introduction Caucasus, east-central Georgia, the southern slopes of the eastern Great Caucasus and south-east Azerbaijan. A In order to halt ongoing biodiversity loss, conservation regression tree analysis Chi squared automatic interaction efforts are often focussed using cross-country conservation detector method revealed the mean temperature of the plans and biodiversity action plans, which are based on driest quarter and precipitation parameters to be the main existing threats to biodiversity in the region and precise environmental factors shaping these patterns. Diversity of information on the distribution of species (Ceballos and endemic species was correlated with overall species Brown 1995; Garcıá 2006; Newbold et al. 2009; Arponen diversity but hotspots of endemic species (10? percent of 2012). Due to a lack of research on arthropod taxa, con- all species) exists in high-mountain areas, suggesting post- servation plans do not normally include arthropod species glacial speciation events in the high mountains as the main to an extent reflecting their outstanding contribution to the sources of high endemism in Caucasus. Further informa- overall species diversity. This in turn leads to conservation tion on the spatial distribution of species diversity of efforts, which do not effectively cover areas important for arthropod diversity (Hernańdez-Manrique et al. 2012) and therefore probably result in a dramatically increased diversity loss within this taxon, and hence overall species & G. Chaladze ( ) diversity. In order to actually halt current diversity loss, it Institute of Ecology, Ilia State University, 3/5 Cholokashvili Ave., 0162 Tbilisi, Georgia is therefore necessary to study patterns of arthropod e-mail: [email protected] diversity more intensively and use the obtained insights in updated conservation plans and biodiversity action plans S. Otto (Cardoso et al. 2008; Diniz-Filho et al. 2010; Beck et al. Department of Animal Ecology and Tropical Biology, University Wu¨rzburg, GutsMuthsstr. 42, 04177 Leipzig, 2012; Hernańdez-Manrique et al. 2012). Germany Because of its importance as one of the worldwide Biodi- versity Hotspots (Myers et al. 2000; Kier et al. 2005; Foster- S. Tramp Turley and Gokhelashivili 2009; Zazanishvili and Mallon Department of Computer Science, Business Information Systems, University Leipzig, Augustusplatz 10, Zimmer P614, 2009), a number of conservation and action plans have been 04109 Leipzig, Germany published within the Caucasus Ecoregion (henceforth termed

123 Author's personal copy

J Insect Conserv

CE) (MEPNR 2005; Williams et al. 2006; MEPNR 2011). taxon. However, a recent study modeled the distribution of Despite high total numbers of species and high rates of ant species richness in Georgia (Chaladze 2012), facilitat- endemism among arthropods (Aliyev et al. 2009; Kalashian ing new hypotheses on the location of arthropod diversity 2009; Konstantinov et al. 2009; Zazanishvili and Mallon hotspots in this country. Next to intensified field work, it is 2009), including spiders (Ysnel et al. 2008), knowledge about important that more studies retrieve the richness of existing spatial patterns of arthropod species diversity in the CE is occurrence data and provide distribution maps of the virtually nonexistent. In order to give the arthropods their diversity of additional arthropod taxa. Using such maps of deserved weight for conservation efforts in this region, it is different arthropod species to create spatial representations important to close this data gap by both intensified research as of the overall arthropod diversity should yield the infor- well as by making the existing data available in freely mation needed to give the highly diverse arthropods their accessible databases. As Otto and Tramp (2012) showed for deserved status in future conservation plans for the CE. spiders in the CE, reviewing the existing literature for a given In the present paper we aim to contribute to this goal by taxon and compiling these occurrence data into a database can providing spatial models of overall and endemic spider dramatically increase the number of known occurrences and species diversity based on SDMs. We think that spiders are update the species lists for the relevant countries in this region a good model taxon for this study because regional and (Mikhailov 2002; Marusik et al. 2006; Ysnel et al. 2008). continental spider diversity patterns can be explained to a A large amount of occurrence data of sufficient quality large extent by environmental factors (Jimeńez-Valverde is often insufficient on its own to derive the information and Lobo 2007; Finch et al. 2008; Jimeńez-Valverde et al. needed for effective species conservation. In order to 2010; Carvalho et al. 2012), commonly included in SDM identify hotspots of arthropod diversity, endemic species or approaches, e.g. climatic and topographic factors. Here we threatened species, the occurrence data must be translated aim to answer the following questions: into spatial models of distribution for every species, (1) What is the predicted spatial pattern of spider species resulting in maps highlighting areas of high arthropod richness in the CE? diversity. Macroecological methods like species distribu- (2) Where are predicted hotspots of spider diversity tion modeling (SDM) can bridge this gap between existing located in the CE? occurrence data and the final distribution maps (Arau´jo and (3) Where are predicted hotspots of endemic spider Peterson 2012). In SDM, species occurrence data can be diversity located in the CE? correlated with abiotic (scenopoetic), biotic and movement (4) Where are regions predicted to show extraordinarily factors (biogeographic and migratory) in the region of high proportions of endemic spider species in the interest, in order calculate a spatial model of the area of CE? distribution with suitable conditions for this species (Syp- (5) What are the underlying factors shaping these hard and Franklin 2009; Graham et al. 2010; Soberoń 2010; patterns? Zimmermann et al. 2010; Vasconcelos et al. 2012). Recent algorithms and statistical methods have helped to develop spatial models describing biodiversity including those developed for the prediction of species distributions Materials and methods (Stockwell and Peters 1999; Soberon and Peterson 2005; Fitzpatrick et al. 2007; Ortega-Huerta and Peterson 2008; Study area de Souza Munõz et al. 2011). In order to identify the spatial pattern of species richness, distribution models for single The study area includes the political territories of Georgia, species are developed and then stacked (Garcıá 2006; Armenia, Azerbaijan as well as the countries of the North Newbold et al. 2009; Chaladze 2012). An alternative Caucasus and the rayons Krasnodar and Stavropol in method is recording species richness at individual localities Southern Russia (Fig. 1). The CE is situated on the and modeling richness patterns directly. Newbold et al. boundary of temperate and moist-temperate climate belts. (2009) compared the two approaches while modeling the Due to the dominance of its mountainous regions, the cli- butterfly and mammal fauna of Egypt. They showed that matic conditions in the CE are very diverse, ranging from using the former approach (summing individual models) warm and moist regions with a precipitation of more than produces more accurate output. Summing individual 2,000 mm per year near the Black Sea Coast to semi-arid models is a good approach only when the available dis- regions in Azerbaijan, receiving only 250 mm annual tribution data are sufficient to create individual species precipitation (see details in Williams et al. (2006). distribution models. Together with this orographic and climatic complexity Species distribution modeling has rarely been applied in the CE is rich in landscape types of a number of terrestrial the CE to visualize the spatial distribution of an arthropod ecosystems: mountain forests, freshwater and marine

123 Author's personal copy

J Insect Conserv

Fig. 1 Study area. Spots indicate unique localities with occurrence data ecosystems, dry mountain shrublands, steppes, semideserts, precipitation seasonality, (16) precipitation of wettest wetlands and high-mountains, which contribute to the quarter, (17) precipitation of driest quarter, (18) precipita- outstanding biodiversity of the CE (Myers et al. 2000; tion of warmest quarter, and (19) precipitation of coldest Williams et al. 2006; Foster-Turley and Gokhelashivili quarter. 2009; Zazanishvili and Mallon 2009). Range models were developed for each species with at least five records (Garcıá 2006). In total, 566 species fit Modeling these criteria. Three-quarters of the occurrence locations were used for training the models and one quarter was used Species occurrence data was taken from the Caucasian for validation. Occurrences were divided randomly as test Spiders Database (Otto and Tramp 2012), a collection of and training points. The accuracy of each model was 11,418 occurrences of 1,078 spider species from 246 lit- assessed using the area under the receiver operator (ROC) erature sources established since 2006. We excluded curve (AUC); the calculations were performed in Open- oversampled locations (Tbilisi and the Lagodekhi National Modeller with supply of test and training occurrences Park) in order to reduce sampling bias (Hortal et al. 2008). independently. Following the recommendations made by The modeling of species distribution was performed using Swets (1986), species with AUC scores of\0.7 were the software package OpenModeller (de Souza Munõz excluded from further analysis. et al. 2011). This software helps to model suitable distribution range for individual species and then overlays them Statistical analysis in order to estimate a summed model of species richness. The GARP algorithm (Stockwell and Peters 1999; Stock- 10,000 random points were generated using Arcview 3.1, well 1999) was used to infer the spiders’ diversity hotspots. covering the whole study area. The following variables In total, 19 variables were taken from the WorldClim were scored for each random point: total inferred species version 1.4 dataset at a resolution of 5 arcmin (c. 10 km) richness, inferred endemic species richness, the 19 biocli- (Hijmans et al. 2005): (1) Annual mean temperature, (2) matic variables listed above and elevation was extracted mean diurnal range, (3) isothermality, (4) temperature from GIS layers using ArcMap 9.3. seasonality, (5) maximum temperature of warmest month, The regression tree analysis with Chi squared automatic (6) minimum temperature of coldest month, (7) tempera- interaction detector [CHAID, Kass (1980)] method was ture annual range, (8) mean temperature of wettest quarter, used in order to determine the interaction between pre- (9) mean temperature of driest quarter, (10) mean tem- dicted species richness and the environmental variables. perature of warmest quarter, (11) mean temperature of CHAID analysis is a non-parametric procedure and no coldest quarter, (12) annual precipitation, (13) precipitation assumptions about the data distribution need to be made of wettest month, (14) precipitation of driest month, (15) (van Diepen and Franses 2006).

123 Author's personal copy

J Insect Conserv

Fig. 2 Distribution of species richness of valid species across the study area. Darker shades indicate higher predicted numbers of species

SPSS software (SPSS v.16) was used to carry out the for predicted species richness: Mean Temperature of Driest analysis. A signiﬁcance level of 5 % was used in the F test, Quarter, Temperature Annual Range, Precipitation of the maximum number of levels was established as three, Coldest Quarter, Temperature Seasonality, Mean Temper- and the minimum number of cases in a node for being a ature of Warmest Quarter, Annual Precipitation, Annual child node was established as 50. Diagrams were compiled Mean Temperature (R2= 0.806 SEE 36.99, Cross-Vali- using SPSS software (SPSS v.16). dation R2= 0.796 SEE= 36.45). Total species richness is best discriminated by the Mean Temperature of Driest Quarter (P\ 0.001) with the highest Results diversity value (Mean 299.6 SD 63.5) predicted where temperature range is between-0.74 and 1.93�C. In those regions All species with the driest quarter of winter below 0�C (Armenian Upland, central Transcaucasus and Great Caucasus range) Of all the 566 species included in the analysis 472 passed species diversity is positively correlated with mean temper- the cross validation test (AUC[ 0.7). Total species rich- ature of driest quarter. At temperatures above 0�C in the ness varied from 0 to 446 with an average value of 196.5. mentioned regions and in the foothills of the study region with Species richness was not distributed uniformly within the a mean temperature of 6–15�C during winter, species study area (Fig. 2). High species richness is predicted for diversity is negatively correlated with temperature. On the regions with mountain forests, especially in the central southeastern Black-Sea Coast and in the Colchis triangle, parts of the North Caucasus, the Transcaucasus between where summer is the driest part of the year, species diversity is the Surami Mountains and Gombori Mountains in East also negatively correlated with temperature in regions with Georgia, the southern slopes of the eastern Great Caucasus mean temperatures above 15�C. as well as in mountains in southwestern Azerbaijan. At high elevations where Mean Temperature of Driest Total Species richness showed a non linear correlation Quarter is below-4.68�C Temperature Annual Range is with elevation, with a well expressed mid-elevation peak second important parameter, species richness being higher (Fig. 3). The lowest richness of endemic species was pre- where Temperature annual range is lower (i.e. more even dicted for very high ([1914 m; Mean 136.9 SD 60.6) and Temperature). Across lowlands of Caspian sea, where Mean low (\157 m a.s.l.; Mean 152.1 SD 53.5) elevations; Temperature of Driest Quarter is higher than 24.19�C Annual highest species richness (Mean 243.3 SD 91.2) was pre- Mean Temperature is second important parameter. Temper- dicted for altitudes between 369 and 1,622 m a.s.l. ature Seasonality is second important parameter on low ele- The regression tree (CHAID algorithm) including 19 vations (200–600 m a.s.l.) and mead elevation climatic variables revealed seven important discriminators (1,100–1,600 m a.s.l.) predicted species richness is negatively

123 Author's personal copy

J Insect Conserv

Fig. 3 Species richness distribution of all species (open circles) and endemic species (ﬁlled circles) by altitude

Fig. 4 Mean temperature of driest quarter, annual precipitation and species richness correlated with temperature seasonality. In rest of areas pre- slopes of the Eastern Great Caucasus as well as parts of the cipitation is second important splitter, highest species richness mountains in southwestern Azerbaijan. (mean 358.8 SD 43.43) predicted where Precipitation of Endemic species richness in this study showed non- Coldest Quarter is 86.85–94.72 mm (Fig. 4). linear correlation with elevation, with a well-expressed mid-elevation peak (Fig. 3). The lowest endemic richness Endemic species was predicted for high elevations ([2,288 m, Mean 9.5, SD 4.8) and low elevations (\369 m, Mean 3.62, SD 3.9). The Endemic species richness varied from 0 to 42 with an highest richness (Mean 15.4, SD 5.53) was predicted for average value of 10.1. Endemic species richness was not elevations between 632 and 1,622 m. distributed uniformly within the study area (Fig. 5). The Endemic species number was correlated with total spe- highest species richness is predicted for the mountains in cies richness (r2 = 0.730,P\ 0.001). The proportion of East Georgia (e.g. the Gombori Mountains), the southern endemic species varied from 0 to 14 %, with average value

123 Author's personal copy

J Insect Conserv

Fig. 5 Distribution of species richness of endemic species across the study area

Fig. 6 Proportion of endemic species at different altitudes of 4.85 %. The proportion of endemic species was posi- Precipitation, Temperature Annual Range, Mean Temper- tively correlated with elevation (R2= 0.758,P\ 0.001). ature of Warmest Quarter (R2 = 0.801, SEE 0.325; Cross- Highes proportions of endemic species (above 10 %) were Validation R2 = 0.786, SEE 0.337). predicted for some regions at very low altitudes at very Endemic species richness, similarly to total species high altitudes, e.g. the Central Greater Caucasus and the richness, is best discriminated by the Mean Temperature of Kars-Armavir Region (Figs. 6, 7). Driest Quarter with the highest diversity value (Mean The regression tree including 19 environmental factors 19.25, SD 6.43) predicted where temperature range is revealed four important discriminators for endemic species between-0.74 and 1.93�C. In those regions with the richness: Mean Temperature of Driest Quarter, Annual Mean Temperature of Driest Quarter below 0�C species

123 Author's personal copy

J Insect Conserv

Fig. 7 Regions with high proportions of endemic species in spiders ([10 %) diversity is positively correlated with the Mean Tempera- confirms the southern slopes of the Eastern Great Caucasus ture of Driest Quarter. At temperatures above 0� predicted as a hotspot of spider diversity, exemplified by the fact that species diversity is negatively correlated with Mean Tem- according to occurrence data Lagdekhi National Park is the perature of Driest Quarter. location with the highest number of spider species (Otto At very low (Mean Temperature of Driest Quarter and Tramp 2012). It was not included in the calculation of [24.1�C) and high elevations (Mean Temperature of our model in order to minimize sampling bias but the Driest Quarter\-4.68�C) Temperature Annual Range is model predicts high diversity in this forest nonetheless. second important parameter, species richness being higher Secondly, locations of known Tertiary refugia, which where Temperature annual range is smaller. Mean Tem- are commonly expected to be hotspots of diversity, perature of Warmest Quarter is second important parameter according to our model are predicted to exhibit only where Mean Temperature of Driest Quarter is between 1.93 intermediate levels of spider diversity, e.g. the Black-Sea and 12.8�C, geographicaly this represent low elevation Coast, the Colchic lowlands and the Hyrkan Forests. It areas between 200 and 600 m.a.s.l. In rest of areas Annual could be that spiders migrated from their refugia to more Precipitation is second important parameter, predicted suitable habitats or climates in other regions as has been richness is positively correlated, with highest value (Mean shown for other taxa (Graham et al. 2010; Zimmermann 22.91, SD 7.28) between 722 and 938 mm of precipitation. et al. 2010), thus increasing diversity outside of the refugia, whereas less motile species became extinct or remained within the refugia (e.g. species of the genus Raveniola Discussion Zonstein, 1987). Vast parts of the predicted hotspots of spider diversity in the central North Caucasus, central and All species eastern Georgia and Southwestern Azerbaijan have so far received little attention in arachnological diversity studies The spatial distribution of spider species richness as pre- (cf. Fig. 1), a data gap in dire need of increased sampling dicted by our model confirms a number of hypotheses efforts in order to improve our knowledge on the distri- found in earlier arachnological studies in the CE. First of bution of spider diversity distribution in the CE. all, spider diversity in temperate and tropical forests is We found the diversity of all species—as well as that of generally high (Sørensen 2004; Cardoso et al. 2008; Otto endemic species—to form the typically hump-shaped dis- and Floren 2010; Blick 2011; Basset et al. 2012); indeed tribution of species-richness-altitude relationships in spi- the locations with the highest reported species numbers are ders and other arthropods (Mikhailov and Mikhailova situated in the forest zone (e.g. Borjomi-Kharagauli 2002; Chatzaki et al. 2005; Werenkraut and Ruggiero National Park, Lagodekhi National Park). Our model also 2010) with highest species numbers at altitudes of the

123 Author's personal copy

J Insect Conserv mountain forest zone. Spider diversity in north-temperate similar size (Ysnel et al. 2008). We found, that the local forests is generally high (Floren et al. 2008; Otto and proportion of endemic species is about 4.8 % (SD 3.10) but Floren 2010; Blick 2011) and can increase with altitude can reach 10–14? % in some regions (Fig. 7). Endemic when conditions are suitable, e.g. in tropical forests (Rus- species richness is positively correlated with elevation, sell-Smith and Stork 1994). As habitat structure and suit- suggesting post-glacial speciation events in the high able climatic conditions dramatically decrease in the higher mountains as the main source of high endemism. alto-montane forests and above the forest zone, a negative The factors influencing overall spider richness and correlation between altitude and spider species richness is endemic species richness are similar: mean temperature of usually observed at high altitudes. Steiner and Thaler the driest quarter is the most important climatic factor and (2004) found such a decrease in species richness in arb- annual precipitation, temperature annual range, mean oricolous spiders at altitudes above 1,000 m in the Euro- temperature of warmest quarter are the second most pean Alps, whereas our model predicts a reduction of important factors. species richness above 1,500 m in the Caucasus (Fig. 3). This difference could possibly be related to the overall Implications for species conservation warmer climate and thus higher-reaching vegetation zones in the Caucasus as compared to the European Alps. In our study we identified the altitude zone of the sub- Some studies found temperature parameters to be good montane to montane forests as harboring the highest spe- predictors of spider abundance and diversity at various cies richness in spiders, a pattern similar to the distribution scales (Rypstra 1986; Jimeńez-Valverde and Lobo 2007; of ant diversity in Georgia (Chaladze 2012). The diversity Finch et al. 2008). This is corroborated by our finding that of endemic species is correlated with total species richness spider diversity is most strongly affected by the mean but exhibits especially high proportions at very high alti- temperature of the driest quarter of the year, depending on tudes in the central parts of the Greater Caucasus. whether this period occurs in winter or summer. In winter, Giving forests in the CE priority in conservation efforts mild temperatures just below zero represent favorable would most likely protect the majority of the arthropod conditions for diapausing spiders, whereas colder temper- species if their distribution patterns prove similar to those atures can decrease survival rates (Scha¨fer 1987; Foelix of ants and spiders in future studies. However, the effective 1996). Our findings that mean winter temperatures just protection of rare and endemic arthropod species needs above freezing are negatively correlated with species more detailed information on the distribution and specific diversity could be attributed to a higher mortality, e.g. due threats to the species in question. to untimely emergence from diapause, repetitive freezing- The most urgent activities for filling these gaps in our thawing events or a combination of raised metabolism and knowledge are the establishment of extensive databases food scarcity in winter (Aitchison 1984; Li and Jackson based on published occurrence data as well as intensified 1996; Bale and Hayward 2010; Schmalhofer 2011). In the field work on all major arthropod taxa. For example, in our Colchis, with the driest quarter in summer, the negative study on spiders, the arthropod taxon with the best data correlation of the mean temperature with species diversity base on occurrence data in the CE, only 471 out of a total might be attributed to the higher risk of desiccation at of 1,078 recorded species could be included in the SDM increased temperature levels (Nentwig 1987; DeVito et al. because of missing data for the remaining species. 2004). On an intermediate temperature level we found Increased data mining and field work is needed to improve annual precipitation to be the second most important pre- SDM in spiders and, even more so, in other arthropod taxa. dictor of spider diversity. Spiders tend to be more abundant in moist habitats (Samu et al. 1996) but a discussion of Acknowledgments We are indebted to David Tarkhnishvili and precipitation is difficult because it does not automatically Alexander Gavashelishvili for productive discussions of species distribution modeling. Jason Dunlop kindly commented our English. reflect the amount of water available (Kerr 2001; Finch et al. 2008). References Endemic species Aitchison CW (1984) Low temperature feeding by winter-active The CE is known for its extraordinarily high rates of spiders. J Arachnol 12(3):297–305 endemic species (Foster-Turley and Gokhelashivili 2009; Aliyev KA, Atakishiyeva AM, Gadjiyeva SA, Huseynzade GA, Zazanishvili and Mallon 2009); in spiders endemism is Huseynov EF, Mammadova TG (2009) Arthropoda of the Hirkan Corridor and Hirkan National Park: Red List Update. In: estimated at approx. 22–23 % (Marusik et al. 2006; Otto Zazanashvili N, Mallon D (eds) Status and Protection of and Tramp 2012), which is the highest rate of endemism in Globally Threatened Species in the Caucasus. Contour Ltd., spiders compared to other west Palearctic regions of CEPF, pp 179–182

123 Author's personal copy

J Insect Conserv

Arau´jo MB, Peterson AT (2012) Uses and misuses of bioclimatic Fitzpatrick MC, Weltzin JF, Sanders NJ, Dunn RR (2007) The envelope modeling. Ecology 93(7):1527–1539. doi:10.1890/11- biogeography of prediction error: why does the introduced range 1930.1 of the fire ant over-predict its native range? Glob Ecol Biogeogr Arponen A (2012) Prioritizing species for conservation planning. 16(1):24–33. doi:10.1111/j.1466-8238.2006.00258.x Biodivers Conserv 21:875–893. doi:10.1007/s10531-012-0242-1 Floren A, Otto S, Linsenmair Karl-Eduard (2008) Do spider Bale JS, Hayward SAL (2010) Insect overwintering in a changing communities in primary forests differ from those in forest- climate. J Exp Biol 213(6):980–994. doi:10.1242/jeb.037911 plantations? In: Schmidl J, Floren A (eds) A canopy-study in the Basset Y, Cizek L, Cueńoud P, Didham RK, Guilhaumon F, Missa O, Białowieza-Forest_ (Poland). Canopy arthropod research in Novotny V, Ødegaard F, Roslin T, Schmidl J, Tishechkin AK, Europe, Bioform Verlag, pp 489–506 Winchester NN, Roubik DW, Aberlenc H-P, Bail J, Barrios H, Foelix RF (1996) Biology of Spiders. Oxford University Press, Georg Bridle JR, Castanõ-Meneses G, Corbara B, Curletti G, Duarte da Thieme Verlag Rocha W, De Bakker D, Delabie JHC, Dejean A, Fagan LL, Foster-Turley P, Gokhelashivili R (2009) Biodiversity Analysis Floren A, Kitching RL, Medianero E, Miller SE, Gama de Update for Georgia. USAID, Washington DC Oliveira E, Orivel J, Pollet M, Rapp M, Ribeiro SP, Roisin Y, Garcıá A (2006) Using ecological niche modelling to identify Schmidt JB, Sørensen L, Leponce M (2012) Arthropod diversity diversity hotspots for the herpetofauna of Pacific lowlands and in a tropical forest. Science 338(6113):1481–1484. doi:10.1126/ adjacent interior valley of Mexico. Biol Conserv 130(1):25–46. science.1226727 doi:10.1016/j.biocon.2005.11.030 Beck J, Ballesteros-Mejia L, Buchmann CM, Dengler J, Fritz SA, Graham CH, VanDerWal J, Phillips SJ, Moritz C, Williams SE (2010) Gruber B, Hof C, Jansen F, Knapp S, Kreft H, Schneider A-K, Dynamic refugia and species persistence: tracking spatial shifts Winter M, Dormann CF (2012) What’s on the horizon for in habitat through time. Ecography 33(6):1062–1069. doi:10. macroecology? Ecography 35(8):673–683. doi:10.1111/j.1600- 1111/j.1600-0587.2010.06430.x 0587.2012.07364.x Hernańdez-Manrique OL, Numa C, Vverdu´ JR, Galante E, Lobo JM Blick (2011) Abundant and rare spiders on tree trunks in German (2012) Current protected sites do not allow the representation of forests (Arachnida, Araneae). Arachnologische Mitt 40:5–14. endangered invertebrates: the Spanish case. Insect Conserv doi:10.5431/aramit4002 Divers 5(6):414–421. doi:10.1111/j.1752-4598.2011.00175.x Cardoso P, Scharff N, Gaspar C, Henriques SS, Carvalho R, Castro Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very PH, Schmidt JB, Silva I, Szu¨ts T, De Castro A, Crespo LC high resolution interpolated climate surfaces for global land (2008) Rapid biodiversity assessment of spiders (Araneae) using areas. Int J Climatol 25(15):1965–1978. doi:10.1002/joc.1276 semi-quantitative sampling: a case study in a Mediterranean Hortal J, Jimeńez-Valverde A, Go´mez JF, Lobo JM, Baselga A (2008) forest. Insect Conserv Divers 1(2):71–84. doi:10.1111/j.1752- Historical bias in biodiversity inventories affects the observed 4598.2007.00008.x environmental niche of the species. Oikos 117:847–858. doi:10. Carvalho JC, Cardoso P, Crespo LC, Henriques S, Carvalho R, 1111/j.2008.0030-1299.16434.x Gomes P (2012) Determinants of spider species richness in Jimeńez-Valverde A, Lobo JM (2007) Determinants of local spider coastal dunes along a gradient of mediterraneity. Insect Conserv (Araneidae and Thomisidae) species richness on a regional scale: Divers 5(2):127–137. doi:10.1111/j.1752-4598.2011.00139.x climate and altitude vs. habitat structure. Ecol Entomol Ceballos G, Brown JH (1995) Global patterns of mammalian 32(1):113–122. doi:10.1111/j.1365-2311.2006.00848.x diversity, endemism, and endangerment. Conserv Biol Jimeńez-Valverde A, Baselga A, Melic A, Txasko N (2010) Climate 9(3):559–568. doi:10.1046/j.1523-1739.1995.09030559.x and regional beta-diversity gradients in spiders: dispersal Chaladze G (2012) Climate-based model of spatial pattern of the capacity has nothing to say? Insect Conserv Divers 3(1):51–60. species richness of ants in Georgia. J Insect Conserv doi:10.1111/j.1752-4598.2009.00067.x 16(5):791–800. doi:10.1007/s10841-012-9464-5 Kalashian MY (2009) Rare Invertebrates in Armenia. In: Zazanashvili Chatzaki M, Lymberakis P, Markakis G, Mylonas M (2005) The N, Mallon D (eds) Status and Protection of Globally Threatened distribution of ground spiders (Araneae, Gnaphosidae) along the Species in the Caucasus. Contour Ltd., CEPF, pp 183–187 altitudinal gradient of Crete, Greece: species richness, activity Kass GV (1980) An exploratory technique for investigating large and altitudinal range. J Biogeogr 32(5):813–831. doi:10.1111/j. quantities of categorical data. Appl Stat 29(2):119–127 1365-2699.2004.01189.x Kerr JT (2001) Butterfly species richness patterns in Canada: energy, Connor SE, Kvavadze EV (2009) Modelling late quaternary changes heterogeneity, and the potential consequences of climate change. in plant distribution, vegetation and climate using pollen data Conserv Ecol 5(1):10 from Georgia Caucasus. J Biogeogr 36(3):529–545. doi:10.1111/ Kier G, Mutke J, Dinerstein E, Ricketts TH, Ku¨per W, Kreft H, j.1365-2699.2008.02019.x Barthlott W (2005) Global patterns of plant diversity and floristic de Souza Munõz M, De Giovanni R, de Siqueira M, Sutton T, Brewer knowledge. J Biogeogr 32(7):1107–1116. doi:10.1111/j.1365- P, Pereira R, Canhos D, Canhos V (2011) openModeller: a 2699.2005.01272.x generic approach to species’ potential distribution modelling. Konstantinov AS, Korotyaev BA, Volkovitsh MG (2009) Insect GeoInformatica 15:111–135. doi:10.1007/s10707-009-0090-7 Biodiversity in the Palearctic Region. In: Foottit RG, Adler PH DeVito J, Meik JM, Gerson MM, Formanowicz DR Jr (2004) (eds) Insect Biodiversity, Wiley-Blackwell, pp 107–162, doi: 10. Physiological tolerances of three sympatric riparian wolf spiders 1002/9781444308211.ch7 (Araneae: Lycosidae) correspond with microhabitat distribu- Leroy SAG, Arpe K (2007) Glacial refugia for summer-green trees in tions. Can J Zool 82(7):1119–1125. doi:10.1139/z04-090 Europe and south-west Asia as proposed by ECHAM3 time-slice Diniz-Filho JAF, De Marco P Jr, Hawkins BA (2010) Defying the atmospheric model simulations. J Biogeogr 34(12):2115–2128. curse of ignorance: perspectives in insect macroecology and doi:10.1111/j.1365-2699.2007.01754.x conservation biogeography. Insect Conserv Divers Li D, Jackson RR (1996) How temperature affects development and 3(3):172–179. doi:10.1111/j.1752-4598.2010.00091.x reproduction in spiders: a review. J Therm Biol 21(4):245–274. Finch O-D, Blick T, Schuldt A (2008) Macroecological patterns of doi:10.1016/0306-4565(96)00009-5 spider species richness across Europe. Biodivers Conserv Marusik Y, Mikhailov KG, Guseinov EF (2006) Advance in the study 17(12):2849–2868 of biodiversity of Caucasian spiders (Araneae). In: Deltshev C,

123 Author's personal copy

J Insect Conserv

Stoev P (eds) European Arachnology 2005, Acta zoologica T (eds) Relict Species, Springer, Berlin, pp 239–252, doi: 10. bulgarica, Suppl. No. 1, pp 259–268 1007/978-3-540-92160-8_13 MEPNR (2005) National Biodiversity Strategy and Action Plan - Soberoń JM (2010) Niche and area of distribution modeling: a Georgia. Ministry of Environment Protection and Natural population ecology perspective. Ecography 33(1):159–167. Resources, Tbilisi, Georgia, p 106, url: www.cbd.int/doc/ doi:10.1111/j.1600-0587.2009.06074.x world/ge/ge-nbsap-01-en.pdf Soberon J, Peterson AT (2005) Interpretation of models of funda- MEPNR (2011) National Environmental Action Plan of Georgia mental ecological niches and species’ distributional areas. 2011–2015. Ministry of Environment Protection and Natural Biodivers Inform 2:1–10 Resources, Tbilisi, Georgia, p 114, url: http://moe.gov.ge/files/ Sørensen LL (2004) Composition and diversity of the spider fauna in Tenderebi/NEAP-2_Final_Eng_6.doc the canopy of a montane forest in Tanzania. Biodivers Conserv Mikhailov KG (2002) The spider fauna of Russia and other post- 13(2):437–452. doi:10.1023/B:BIOC.0000006510.49496.1e Soviet republics: a 2000 update. In: Toft S, Scharff N (eds) Steiner E, Thaler K (2004) Ho¨henverteilung arborikoler Spinnen European Arachnology 2000 (Proceedings of the 19th European (Arachnida: Araneae) im Gebirgswald der Zentralalpen (Pat- Colloquium of Arachnology). Aarhus University Press, Aarhus, scherkofel bei Innsbruck, Nordtirol). Berichte des naturwis- pp 255–259 senschaftlich-medizinischen Vererins in Innsbruck, 91:157–185 Mikhailov KG, Mikhailova EA (2002). Altitudinal and biotopic Stockwell DRB (1999) Genetic algorithms II. In: Fielding AH (ed) distribution of the spider family Gnaphosidae in North Ossetia Machine learning methods for ecological applications, Kluwer, (Caucasus major), pp 261–265 Boston, pp 123–144 Muster C, Berendonk TU (2006) Divergence and diversity: lessons Stockwell D, Peters D (1999) The GARP modelling system: problems from an arctic–alpine distribution (Pardosa saltuaria group, and solutions to automated spatial prediction. Int J Geogr Inf Sci Lycosidae). Mol Ecol 15(10):2921–2933. doi:10.1111/j.1365- 13(2):143–158. doi:10.1080/136588199241391 294X.2006.02989.x Swets JA (1986) Indices of discrimination or diagnostic accuracy: Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J their ROCs and implied models. Psychol Bull 99(1):100–117. (2000) Biodiversity hotspots for conservation priorities. Nature doi:10.1037/0033-2909.99.1.100 403(6772):853–858 Syphard AD, Franklin J (2009) Differences in spatial predictions Nentwig W (ed) (1987) Ecophysiology of spiders. Springer Verlag, among species distribution modeling methods vary with species Berlin Heidelberg traits and environmental predictors. Ecography 32(6):907–918. Newbold T, Gilbert F, Zalat S, El-Gabbas A, Reader T (2009) doi:10.1111/j.1600-0587.2009.05883.x Climate-based models of spatial patterns of species richness in Tarkhnishvili D, Gavashelishvili A, Mumladze L (2012) Palaeocli- Egypt’s butterfly and mammal fauna. J Biogeogr matic models help to understand current distribution of Cauca- 36(11):2085–2095. doi:10.1111/j.1365-2699.2009.02140.x sian forest species. Biol J Linn Soc 105(1):231–248. doi:10. Ortega-Huerta MA, Peterson AT (2008) Modeling ecological niches 1111/j.1095-8312.2011.01788.x and predicting geographic distributions—a test of six presence- R Development Core Team, (2012) R: A Language and Environment only methods. Rev Mex Biodivers 79:205–216 for Statistical Computing, R Foundation for Statistical Comput- Otto S, Floren A (2010) The canopy spiders (Araneae) of the ing, Vienna, Austria, ISBN 3-900051-07-0, url: http://www.R- floodplain forest in Leipzig. Arachnologische Mitt 39:25–38. project.org doi:10.5431/aramit3904 van Diepen M, Franses PH (2006) Evaluating Chi squared automatic Otto S, Tramp S (2012) Caucasian Spiders—A Faunistic Database on interaction detection. Info Syst 31(8):814–831. doi:10.1016/j.is. the Spiders of the Caucasus, url: http://db.caucasus-spiders.info 2005.03.002 Accessed in January Vasconcelos TS, Rodrı´guez M, Hawkins BA (2012) Species distri- Russell-Smith A, Stork NE (1994) Abundance and diversity of spiders bution modelling as a macroecological tool: a case study using from the canopy of tropical rainforests with particular reference New World amphibians. Ecography 35(6):539–548. doi:10. to Sulawesi, Indonesia. J Trop Ecol 10:545–558. doi:10.1017/ 1111/j.1600-0587.2011.07050.x S0266467400008221 Werenkraut V, Ruggiero A (2010) Quality of basic data and method Rypstra AL (1986) Web spiders in temperate and tropical forests: to identify shape affect richness–altitude relationships in meta- relative Abundance and environmental correlates. Am Midl Nat analysis. Ecology 92(1):253–260. doi:10.1890/09-2405.1 115:42–51 Williams L, Zazanashvili N, Sanadiradze G, Kandaurov A (2006) An Samu F, Sunderland KD, Topping CJ, Fenlon JS (1996) A spider Ecoregional Conservation Plan for the Caucasus. WWF Cauca- population in flux: selection and abandonment of artificial web- sus Programme Office, Tbilisi sites and the importance of intraspecific interactions in Lepthy- Ysnel F, Pe´tillon EG, Canard A (2008) Assessing the conservation phantes tenuis (Araneae: Linyphiidae) in wheat. Oecologia value of the spider fauna across the West Palearctic area. 106:228–239 J Arachnol 36(2):457–463 Scha¨fer M (1987) Life Cycles and Diapause. In: Nentwig, W. (ed) Zazanishvili N, Mallon D (2009) Status and Protection of globally Ecophysiology of Spiders, Springer, pp 331–347 threatened species in the Caucasus—CEPF Biodiversity Invest- Schmalhofer VR (2011) Impacts of temperature, hunger and repro- ments in the Caucasus Hotspot 2004–2009. CEPF, WWF, ductive condition on metabolic rates of flower-dwelling crab Contour Ltd., Tbilisi, Georgia spiders (Araneae: Thomisidae). J Arachnol 39(1):41–52 Zimmermann NE, Edwards TC, Graham CH, Pearman PB, Svenning Schmitt T, Muster C, Scho¨nswetter P (2010) Are Disjunct Alpine and J-C (2010) New trends in species distribution modelling. Arctic-Alpine Animal and Plant Species in the Western Ecography 33(6):985–989. doi:10.1111/j.1600-0587.2010. Palearctic Really ‘‘Relics of a Cold Past’’? In: Habel J, Assmann 06953.x

123